The Basics of Blocklists
Blocklists are one way that a Mastodon instance can handle unwanted content on the instance level. When a Mastodon domain is on the Blocklist, this means that the server administrators have limited or completely suspended activity with the server at that domain. The specific actions are:
This is the most commonly known one, as it prevents all instance-to-instance activity.
Previously known as “silencing”. This means that accounts on the home server can follow accounts on the limited server, but that general content from that server does not show up on the Federated timeline.
- Reject Media
Media from the moderated server will not display on the home server, this includes not only media in the posts themselves but avatars and headers as well.
Moderator vs User Actions
Users can also take individual action to prevent themselves from seeing unwanted content by blocking or limiting other accounts on the user level. This is recommended for cases where the content is perhaps unwanted by an individual user, but that content does not violate the home server’s ethos.
Mastodon has additional documentation about actions that individual users can take on their Dealing With Unwanted Content documentation page.
How Hachyderm’s Blocklist is Built
At Hachyderm, we do our best to balance what actions should be taken at the instance level and what should be handled at the user level - both when it comes to our own users and when we receive reports of users on other instances or the instances themselves. The vast majority of the moderator action we have taken on servers on our Blocklist is to either silence/limit or suspend/block. The domains that are included on this list are:
- Curated from a variety of published Blocklists on other Mastodon instances
- Added based on user reports. This can mean either:
- A domain has been reported, researched, and moderated individually, or
- A high volume of reports regarding several users on an instance have led to an instance being researched and moderated
Although suspension is the most well-known and discussed moderation action, domains may be limited as well. For example, we might limit a server that has several bots that are taking over the Federated timeline but not suspend so users can continue to follow individual accounts.
Concerns that go into building the Blocklist
There are a few top level concerns that go into determining adding servers to the Blocklist:
- Does the ethos of a specific server violate the ethos of this server?
A few easily understood examples would be if a server is anti-Black, anti-queer, or endorses either direct or dog-whistled hate speech and content.
- Is there a security concern around this server?
As a broad example, if we see malicious traffic coming from a specific server or servers.
- What risk does the server pose to our users?
We prioritize the safety and experience of our users that are in historically underrepresented groups.
- What do our user reports look like? Are they user level on a given server or are they indicative of malicious patterns server-wide?
- What level of moderation is needed? Limit/silence or block/suspend?
It is our opinion that it is in our users’ best interest to federate with as much of the Fediverse as possible so that we can all share our joys, sorrows, growth, learning, etc. with each other.
Our goal with the maintenance of the Blocklist is to ensure that all of our users are safe on Hachyderm. That means when we move forward with taking moderation action on a server, that we will take the best course of action to ensure that safety. We will prioritize the safety of our marginalized users over the broader experience of a completely open and unmoderated (at the server level) Fediverse.
When we take moderation action against a server, we consider:
- The items called out in the numbered list above
- Balancing open participation with curating a safe space
Concerns that go into transparency around the Blocklist
Some servers have reasons attached to their moderation action and others do not. In addition, we may or may not announce when we limit/silence or block/suspend individual instances. Why is this?
When we choose what level of notification to send, and how transparent to be, with moderation actions we consider:
- The impact of the change
- The interest level in the change
- The risk of publicizing/being transparent about the change
For example, if we were to take moderation action against any of the large, popular Mastodon instances we would err on the side of transparency as this would be a significant and user impacting event.
Whenever we take action on a server that has malicious activity, and this can be in the form of attacks on our server or in the form of social attacks like stalking and harassment, we err on the side of safety. This means what level of information we provide and how loudly (notifications, etc.) we provide it will be based on what is safest for all of our users.
The vast, vast majority of instances fall between these two extremes and thus the resulting decisions do not fall perfectly in the “fully transparent” or “completely silent” buckets. This means:
- Servers may not have reasons attached to their moderation
decisions, but that doesn’t mean they are a security or safety concern.
To do this would immediately out them as a security concern, by the absence of information.
- Not all changes to individual server status will be loudly announced, but some
We will use the decision-making process outlined above when announcing.
- Not announcing that a server has changed status does not mean that server is or was a cause of concern. It can also, and frequently will, mean that the server size doesn’t impact enough of our users to make a large announcement.
Requesting Moderation Changes for a Server
What to do if there is a domain on the Blocklist in error
We are all human and are prone to mistakes. If there is a domain that is moderated on our Blocklist that seems to be in error, please open a GitHub Issue in our Community repo to request that we take another look at the domain. Please include as much relevant context as you can to help us make our decision. Note that depending on the circumstances, and as outlined above, we may not be able to be fully transparent with our decision - but we commit to erring on the side of transparency with these reports as often as possible. For more information about how to file a report in our community repo, please take a look at our Reporting Documentation.
What to do if you would like us to moderate a server
If there is a server that is not currently moderated, i.e. either limited/silenced or banned/suspended, then please file a report via the Hachyderm (Mastodon) UI or GitHub Issue in our Community repo for us to take a look at that domain or domains. As before, please include as much context as possible. If there is a concern around the domain(s) you would like to report that would be risky to report in our GitHub Issue tracker, please email us at firstname.lastname@example.org. For more information about filing reports and how to choose between the Mastodon UI and the GitHub Issue tracker, please look at our Reporting Documentation.
What not to do in either of these cases
There are far, far more Hachydermians than moderators. We do not follow tags, posts, etc. to make changes to our Blocklist - we only use the sources outlined above.